I've been using Cursor with Sonnet 3.7 for a while. I'm curious about Claude Code and how it compares. Since I usually work in the terminal, I'm interested in how the development experience differs between them. Does Claude Code support the full 200k context window? I have a lot of questions and not much time to experiment, so I'd appreciate any insights before I consider upgrading my plan.
2 Answers
I've been using Claude Code alongside Cursor and honestly, I think it performs better. It has a much stronger grasp of the code base and manages context more effectively, which saves time on debugging. I’ve switched from Cursor to Claude and it’s definitely worth the test run.
It's tough to give a definitive answer because these tools aren't direct copies. Claude Code does excel in terminal work, thanks to the tweaks by Anthropic, offering a nice intangible advantage. On the flip side, Sonnet 3.7 can tend to over-engineer things, making it a bit more complex. If you're keen to try Claude, consider using the Anthropic API first; it won't take long and is a good way to see if you like it without committing to a full plan.
That's interesting! I've been using Claude 3.7 MAX in Cursor but found some limitations. Do you think Claude Code handles the context more efficiently?